This invention relates to microfabrication, and more particularly to systems, methods, and techniques for manufacturing integrated circuit chips using photon lithography. Still more particularly, the present invention relates to systems, methods, and techniques in connection with programmable masks for microlithography.
Lithography is used to transfer a specific pattern onto a surface. Lithography can be used to transfer a variety of patterns including, for example, painting, printing, and the like. More recently, lithographic techniques have become widespread for use in xe2x80x9cmicrofabricationxe2x80x9dxe2x80x94a major example of which is the manufacture of integrated circuits such as computer chips.
In a typical microfabrication operation, lithography is used to define patterns for miniature electrical circuits. The lithography defines a pattern specifying the location of metal, insulators, doped regions, and other features of a circuit printed on a silicon wafer or other substrate. The resulting semiconductor circuit can perform any of a number of different functions.
Improvements in lithography have been mainly responsible for the explosive growth of computers in particular and the semiconductor industry in general. The major improvements in lithography can, for the most part, be put into two categories: an increase in chip size, and a decrease in the minimum feature size (improvement in resolution). Both of these improvements allow an increase in the number of transistors on a single chip (and in the speed at which these transistors can operate). For example, the computer circuitry that would have filled an entire room in 1960""s technology can now be placed on a silicon xe2x80x9cdiexe2x80x9d the size of your thumbnail. A device the size of a wristwatch can contain more computing power than the largest computers of several decades ago.
One type of lithography that is commonly used in the mass production of computer chips is known as xe2x80x9cparallel lithographyxe2x80x9d. Parallel lithography generally prints an entire pattern at one time. This is usually accomplished by projecting photons through a mask onto a photoresist-coated semiconductor wafer, as shown in FIG. 1. A mask (designated by an xe2x80x9cMxe2x80x9d in FIG. 1) provides a template of the desired circuit. A photoresist coat, which may be a thin layer of material coated on the wafer which changes its chemical properties when impinged upon by light, is used to translate or transfer the mask template onto the semiconductor wafer. In more detail, mask M allows photons (incident light, designated by an xe2x80x9cIxe2x80x9d) to pass through the areas defining the features but not through other areas. An example of a typical mask construction would be deposits of metal on a glass substrate. In a way analogous to the way light coming through a photographic negative exposes photographic paper, light coming through the mask exposes the photoresist. The exposed photoresist bearing the pattern selectively xe2x80x9cresistsxe2x80x9d a further process (e.g., etching with acid, bombardment with various particles, deposition of a metallic or other layer, etc.) Thus, this lithography technique using photoresist can be used to effectively translate the pattern defined by the mask into a structural pattern on the semiconductor wafer. By repeating this technique several times on the same wafer using different masks, it is possible to build multi-layered semiconductor structures (e.g., transistors) and associated interconnecting electrical circuits.
Parallel lithography as described above has the advantage that it is possible to achieve a high throughput since the whole image is formed at once. This makes parallel lithography useful for mass production. However, parallel lithography has the disadvantage that a new mask is required each time one desires to change patterns. Because masks can have very complex patterns, masks are quite costly and susceptible to damage.
For mass production, parallel lithography is usually done using a machine known as a xe2x80x9cstepper.xe2x80x9d As schematically depicted in FIG. 1, a stepper consists of a light source (xe2x80x9cLxe2x80x9d), a place to hold a mask (xe2x80x9cWxe2x80x9d), an optical system (xe2x80x9clensesxe2x80x9d, xe2x80x9cLxe2x80x9d) for projecting and demagnifying the image of the mask onto a photoresist-coated wafer (xe2x80x9cWxe2x80x9d), and a stage (xe2x80x9cSxe2x80x9d) to move the wafer. The process of exposing a wafer using a stepper is summarized in FIG. 2A, and is depicted from a side view in FIGS. 2B-2E. In each exposure, a stepper only exposes a small part of the wafer, generally the size of one chip. Since there are often many separate chips on each wafer, the wafer must be exposed many times. The stepper exposes the first chip (FIG. 2B), then moves (xe2x80x9cstepsxe2x80x9d) over (FIG. 2C) to expose the next chip (FIG. 2D) and repeats this process (FIG. 2E) until the entire wafer is exposed. This process is known as xe2x80x9cstep and repeatxe2x80x9d and is the origin of the name xe2x80x9cstepper.xe2x80x9d
A stepper must also be capable of precisely positioning the wafer relative to the mask. This precise positioning (overlay accuracy) is needed because each lithography step must line up with the previous layer of lithography. A stepper spends a significant portion of its time positioning the stage and the rest exposing the photoresist. Despite the great precision necessary, steppers must be capable of high throughput to be useful for mass production. There are steppers that can process one hundred 8-inch wafers per hour.
One way to increase the usefulness of a chip is to increase its size. In the xe2x80x9cstep and repeatxe2x80x9d example described above, the size of the chip is limited to the exposure size of the stepper. The exposure size is small (roughly 20 mmxc3x9740 mm) because of the cost of an optical system that is capable of projecting a high quality image of the mask onto the wafer. It is very expensive to increase the size of a chip by increasing the exposure size of the stepper (for example, this would require a larger lensxe2x80x94which by itself can cost many hundreds of thousands of dollars). Another approach is to modify a stepper so that light only shines on a subsection of the mask at a given time. Then, the mask and wafer can be scanned (moved relative to the fixed light source) simultaneously until the entire mask is imaged onto the wafer, as in FIGS. 3A-3C. This modified stepper is known as a xe2x80x9cscannerxe2x80x9d or xe2x80x9cscanner/stepperxe2x80x9d.
Scanners offer increased chip size at the expense of increased complexity and mask costs. Because scanner masks are larger, the masks are more fragile and are more likely to contain a defect. The increased size and fragility of the mask mean that the masks for a scanner will be more expensive than the masks for a stepper. Also, because the image is being demagnified, the mask and wafer must be scanned at different speeds, as depicted by the length of the arrows in FIGS. 3A-3C. Because of the great precision required, differential scanning increases the cost and complexity of a scanner when compared with a stepper.
Many chip manufacturers are looking toward future improvements in resolution and/or exposure size to help continue the growth that has driven the semiconductor industry for the past thirty years. The improvements in these areas have been partly the result of improvements in the optical systems used to demagnify the mask and of the use of shorter wavelength light. In particular, modem lithography systems used for mass production are xe2x80x9cdiffraction limitedxe2x80x9d, meaning that the smallest feature size that it is possible to print is determined by the diffraction of light and not by the size of features on the mask. In order to improve the resolution, one must use either a shorter wavelength of light or another technique such as optical proximity correction or phase shifting.
Another option for improving resolution is to put the mask in contact with the wafer, as in FIG. 4; the effects of diffraction can be lessened by not giving the light a chance to xe2x80x9cspread outxe2x80x9d after it passes through the mask. Unfortunately, contact lithography is not suitable for mass production for at least two reasons. First, the mask must now be the same size as the final pattern, making the mask more expensive and more fragile. Second, because the mask is in contact with the wafer, it is easily damaged.
The present invention overcomes many of the disadvantages of prior lithographic microfabrication processes while providing further improvements that can significantly enhance the ability to make more complicated semiconductor chips at lower cost.
One aspect provided by this invention provides a new type of programmable structure for exposing a wafer. The programmable structure allows the lithographic pattern to be changed under electronic control. This provides great flexibility, increasing the throughput and decreasing the cost of chip manufacture and providing numerous other advantages.
The programmable structure provided in accordance with one example embodiment of the invention consists of an array of shutters that can be programmed to either transmit light to the wafer (referred to as its xe2x80x9copenxe2x80x9d state) or not transmit light to the wafer (referred to as its xe2x80x9cclosedxe2x80x9d state). A simplified example lithography system incorporating such a programmable mask is schematically depicted in FIG. 5A exposing an example pattern. In FIG. 5B the same programmable mask PPM is shown exposing a different pattern.
The programmable mask shown in FIGS. 5A and 5B can provide a two-dimensional array of individual shutters each of which can be programmed to either transmit light (xe2x80x9copenxe2x80x9d, xe2x80x9ctransparentxe2x80x9d) or block light (xe2x80x9cclosedxe2x80x9d, xe2x80x9copaquexe2x80x9d). At least one such two-dimensional array of structures can be placed between a wafer and a source of electromagnetic energy. Each of the structures may comprise an active region supporting an electron distribution that can be changed to affect the modulation of electromagnetic energy from said source. The structures can be controlled to selectively modulate, in accordance with a programmable pattern, electromagnetic energy impinging on the wafer.
In accordance with this aspect provided by the invention, a system for exposing a wafer may comprise a source of electromagnetic energy, a collimating lens optically coupled to the electromagnetic energy source, a wafer stage, and a two-dimensional array of structures disposed between the wafer stage and the collimating lens. Each of the structures in the array may comprise an active region supporting an electron distribution that can be changed to affect the modulation of electromagnetic energy from said source. An electrical controller coupled to the two-dimensional array may be used to electrically control the semiconductor structures to selectively modulate, in accordance with a changeable pattern, electromagnetic energy from the source that is directed toward the wafer stage.
In accordance with a further aspect provided by the present invention, the programmable structure can comprise or include an array of selective amplifiers. In accordance with this aspect provided by the invention, a programmable electromagnetic energy modulating structure comprises a two-dimensional array of solid-state selective amplifiers each comprising regions of permanently opaque material and active regions. Control circuitry disposed within the array can be provided to selectively control each of the active regions to toggle between an amplifying state and a non-amplifying state. Thus, each selective amplifier is programmed to either amplify light (somewhat analogous to the xe2x80x9copenxe2x80x9d or xe2x80x9ctransparentxe2x80x9d state of a shutter) or be xe2x80x9cnon-amplifyingxe2x80x9d(its xe2x80x9cclosedxe2x80x9d or xe2x80x9copaquexe2x80x9d state). In the non-amplifying state, some portion of the incident light is transmitted through the amplifier material. The portion incident light that is transmitted through the amplifier can range from 0-100%, depending on the specific design and operating conditions. Selective amplification has all of the advantages of a programmable structure that uses shutters with several added advantagesxe2x80x94including reduction in the time required to expose the resist.
In accordance with a further aspect provided by the invention the shutters and selective amplifiers can work in tandem to form a xe2x80x9cprogrammable layerxe2x80x9d. When the programmed pattern calls for light to pass (or, not pass) through a particular pixel, both the selective amplifier and shutter corresponding to that pixel would be put into their open (or, closed) state. FIGS. 6A-6F schematically depicts the operation of an example shutter (labeled as xe2x80x9cSHxe2x80x9d) (FIGS. 6A-6B), an example selective amplifier (labeled as xe2x80x9cAMxe2x80x9d) (FIGS. 6C-6D), and an example device (labeled as xe2x80x9cXxe2x80x9d) combining the two (FIGS. 6E-6F). In each of these figures, xe2x80x9cIxe2x80x9d represents the intensity of light incident on the shutter/amplifier, and xe2x80x9cIxe2x80x2xe2x80x9d represents the light intensity after interacting with the shutter/amplifier. Combining a selective amplifier with a shutter in this manner achieves increased contrast over selective amplification alone.
In accordance with another aspect provided by the present invention, a programmable technique is provided for creating a pattern to be imaged onto a wafer that can be implemented as a viable production technique. Thus, the present invention also provides a technique of making integrated circuits. In accordance with this aspect provided by the invention, a wafer having a surface covered with photoresist is placed on a movable wafer stage. A source directs electromagnetic energy toward a two-dimensional array of semiconductor structures disposed between the source and the wafer stage. The electron distribution within the structures is electrically controlled to defame a desired microfabrication exposure pattern that modulates electromagnetic energy from said source that impinges on the wafer in accordance with a pattern. The modulated energy is used to expose the photoresist with the pattern. The wafer is then etched to selectively remove portions of the photoresist based on the desired microfabrication exposure pattern, and the etched wafer is treated to construct a semiconductor structure layer on the wafer.
In accordance with a further aspect provided by this invention, a diffraction limiter can be used to provide certain advantages associated with contact lithography without requiring some of the disadvantages of contact lithography. In accordance with this aspect of the invention, the diffraction limiter may provide an opaque layer in which there is an array of transparent regions (xe2x80x9cholesxe2x80x9d) distributed in a one-to-one correspondence to the selective amplifiers/shutters. The diffraction limiter is placed in contact with the wafer, and the light that passes through the programmable layer is incident upon it. The diffraction limiter allows the advantages of contact lithography while maintaining the distance between the programmable layer and wafer.
In accordance with a farther aspect provided by the present invention, a programmable shutter array, a programmable selective amplifier array, and a diffraction limiter can all be used in a common system. For example, FIG. 7A depicts schematically a lithography setup incorporating the above three components. FIG. 7B shows a zoomed-in view of the diffraction limiter (denoted by xe2x80x9cDxe2x80x9d) and wafer section. These three components provide a programmable lithography system that offers high throughput, extremely accurate pattern reproduction, and excellent resolution.
Implementing a diffraction limiter in conjunction with any programmable lithography system should significantly reduce the disadvantages associated with contact lithography. This device is placed in contact with the wafer and thus reduces the effects of diffraction (see FIGS. 7A and AB.) Even though the diffraction limiter is close to the wafer and could be damaged, it is inexpensive and is easily replaced. Additionally, one can apply techniques such as phase shifting and/or optical proximity correction on the diffraction limiter itself. Because of the simple, regular shape of the pixels on the diffraction limiter, such corrections should be easily optimized.
Lithography in accordance with the present invention potentially allows a high throughput to be achieved. For example, only a single programmable structure is necessary to print any desired pattern. A non-exhaustive list of some of the many features and advantages provided by the present invention are as follows:
Programmable lithography offers increased flexibility over conventional parallel lithography. This increased flexibility means that a greater variety of chips can be easily produced. It also opens up ways to improve the manufacture of all types of semiconductor products. It also simplifies the entire process of designing and manufacturing semiconductor products.
No need to have different masks to produce different chips. Because a single programmable structure can be programmed with an arbitrary pattern, it is no longer necessary to fabricate (and purchase) new mask sets in order to print new chips. This is extremely cost-effective for producing small quantities of specialized chips because the cost of a mask set can be prohibitively expensive. In producing large quantities of chips the cost of the mask set is less significant because it is a fixed cost amortized over many more chips.
Electronic alignment. Because it is extremely important to line up the current lithography step with previous steps, steppers spend a significant amount of time mechanically aligning the wafer with the mask. With programmable lithography the pattern can be programmed into the xe2x80x9cprogrammable layerxe2x80x9d (i.e. the programmable area within the mask) such that it is aligned with the wafer. This saves time over having to mechanically align as in the case of a conventional mask.
Disconnecting the size of the chip from the exposure size of the stepper. In conventional parallel lithography, the size of a chip is determined by the exposure size of the stepper. However, with programmable lithography a different pattern can be loaded into the programmable layer at each exposure. Hence, there is no longer any reason that the same pattern needs to be imaged each time the stepper does an exposure. Consequently, in programmable lithography, the size of the chip is not intrinsically determined by the size of each individual exposure. This xe2x80x9cdisconnectxe2x80x9d between chip and exposure size is a significant advantage of programmable lithography because chip performance and chip size are closely tied together.
Simplified optical proximity correction. Optical proximity correction is a technique that is used to increase the resolution of the optical system at a given wavelength of light. Because of diffraction, the pattern on the mask is not faithfully reproduced on the wafer. In order to compensate for this, the pattern on the mask can be altered to account for diffraction so that the desired pattern can be imaged onto the wafer. One problem with this technique in a conventional mask is that it is difficult to decide how to alter the shape on the mask such that the desired shape will appear on the wafer. This is mainly due to the size and complexity of the desired pattern. Programmable lithography greatly simplifies this problem because in programmable lithography the same shape (e.g., a square) is always being imaged onto the wafer by each pixel, and the correction can be made on a pixel-by-pixel basis.
Simplified phase shifting. As with optical proximity correction, phase shifting is also a technique that is used to increase the resolution of the optical system at a given wavelength of light. In this technique, a material that causes a phase shift in light is placed on the mask. The phase shifting material causes destructive interference at the wafer between light from neighboring features in order to eliminate the diffraction tails. Phase shifting also suffers from the same problem as optical proximity correction; due to the size and complexity of the pattern it is difficult to decide where to place the phase shifting material. As with optical proximity correction, programmable lithography allows this problem to be greatly simplified because the same shape is always being imaged onto the wafer at each pixel. Programmable lithography also allows the possibility of active phase shifting. In active phase shifting, each pixel would contain an additional layer in which there would be a material whose index of refraction changes when a voltage is applied.
Simplification of the chip making process. One of the big problems facing chip manufacturers is the growing complexity of the chip making process. With each new chip the manufacturer must get a brand new mask set. They must also inspect and repair these masks. With programmable lithography only a single programmable structure is needed to produce an entire chip. In the event that a programmable mask breaks it can simply be replaced with another identical programmable mask. Additionally, programmable lithography will facilitate research and development of new products because of the greater ease of producing prototype devices.
The shutters and/or selective amplifiers can be fabricated easily. Each individual shutter can be a device similar to devices that are typically used in chips themselves. This allows the design and fabrication of the shutter to draw on the enormous amount of knowledge associated with production techniques and operation of these devices.
The shutters and/or selective amplifiers can be small and densely packed. Small shutters mean that the lithography system can produce small features without the need for demagnification, although it could be used in a system that does include demagnification. Densely packed shutters mean high throughput because they reduce the number of exposures necessary to expose the entire wafer.
The shutters and/or selective amplifiers can work for short wavelength light. Shorter wavelengths provide better resolution, which is important for chip performance.
The shutters and/or selective amplifiers Generally will not break during normal operation. This is important because the mask must be able to flawlessly reproduce the desired image. If even a single pixel is incorrect then the entire chip is likely to be worthless.
The shutters and/or selective amplifiers can switch states quickly. The speed of the shutters is relevant for throughput and may become significant when many shutters must be addressed.
Selective amplifiers can be used alone and/or in combination with programmable shutters. An array of selective amplifiers can be used in the place of or in addition to a PPM to project more light onto the wafer in some areas than in others corresponding to the pattern to be imaged. Or, an array of selective amplifiers can be used in a stand-alone system, e.g., when the non-amplified light is not sufficient to expose the resist and the amplified light is.
Programmable lithography provides a resolution and throughput comparable to conventional parallel lithography while retaining all of the advantages of programmable lithography.